1. Field of the Invention
The present invention relates in general to a system and method for identifying information and more particularly, to a system and method for monitoring and analyzing communications.
2. Description of the Related Art
Advances in technology have led to the widespread practice of exchanging information and electronic interaction via computer networks such as the World Wide Web (WWW) or the Internet. However, with this improved method of communication has come a heightened risk for inadvertently transmitting sensitive information. For example, the proliferation of email-based interaction has increased the risk of employees transmitting sensitive information to unauthorized persons. Similarly, real-time communication in the context of Internet chat rooms and electronic message boards has increased the risk of individuals divulging information about themselves or others that they may not have otherwise disclosed in a different setting.
This increased risk of inadvertently divulging sensitive information is particularly pronounced in the context of children interacting online with individuals they do not know. While the Internet has provided a powerful medium by which children are able to meet and communicate with other children having common interests, online chat rooms and the like have also served as havens for those with more sinister intentions.
The methods currently available to combat these issues are primitive, often requiring individuals to physically read through text which has been previously transmitted and captured in a database. In the employee email context, employers often resort to physically scanning their employees' email files for evidence of unauthorized information transfers. More sophisticated chat room monitoring systems flag keywords and store their occurrences in a database for later review. However, administration of such systems is extremely cumbersome and their effectiveness has been limited. This use of keyword flags to identify abnormal communications is simply ineffectual to properly evaluate the normality of a given conversation since the context in which the keywords are used is ignored. Rather, to obtain a true value of the normality of a conversation, the context in which the keywords are used must be considered and given an appropriate weighting.
Thus, what is needed is a system and method which overcomes the aforementioned problems by providing a reliable and context-sensitive system for monitoring text-based communications.